empty vcf from vg call

Question

[philgev2]

Hi

I am getting an empty vcf file when I run below commands

vg construct -t 10 -r /data/public-data/hg38/hg38_v2_Homo_sapiens_assembly38.fa -v /data/public-data/1000genomes/ALL.wgs.shapeit2_integrated_snvindels_v2a.GRCh38.27022019.sites_chr_1:Y.vcf >1000g.vg

vg prune -k 12 -e 2 -t 16 -p 1000g.vg >1000g.pruned.vg

#### store the graph in the xg/gcsa index pair
vg index -t 10 -p -b /data/users/philge/tmp/ -x 1000g.xg -g 1000g.gcsa -k 16 1000g.pruned.vg

####align a read to the indexed version of the graph
vg map -t 10 -b /data/users/philge/genomegraph/vg/dbsnp/sorted_A019.bam -x 1000g.xg -g 1000g.gcsa > ./alignment/sorted_A019_1000g.gam

####Compute the read support from the gam (with ignoring qualitiy < 5)
vg pack -t 10 -x 1000g.xg -g ./alignment/sorted_A019_1000g.gam -Q 5 -o ./alignment/sorted_A019_1000g.pack

####Generate a VCF from the support
vg call -t 10 -r ./1000g.snarls ./1000g.xg -k ./alignment/sorted_A019_1000g.pack > ./alignment/sorted_A019_1000g_no_aug.vcf

All my input files for vg call contain data. But output vcf is empty. Any suggestions on what could be the reason?

Thanks
Philge

[philgev2]

any suggestions, please? I am stuck with the analysis.

Answer 1

[glenn.hickey]

The commands look okay (though vg snarls seems to be missing). Assuming your snarls file is fine, the other reason for an empty VCF could be not enough mappings found in the pack file. You can use vg depth -k -b 10000 to verify the coverage in your pack file (and with -g to verify the coverage in your GAM).

Answer 2

[kingcohn]

I have a similar issue, did you resolve this?